Summary. Sensitivity to insulin in vivo was studied in six Type 1 diabetic patients without residual insulin secretion and without clinical insulin resistance, and in eight non-diabetic subjects, using the euglycaemic insulin clamp technique. Insulin was infused for four periods of 2 h sequentially at 0.5, 1.0, 2.0 and 5.0 mU. kg -1. min-1; for each insulin infusion period the steady-state plasma free insulin levels were comparable in the diabetic and non-diabetic subjects. The mean + SEM plasma glucose concentration was 4.9 + 0.03 mmol/l in the diabetic subjects (coefficient of variation of plasma glucose values: 5.7 + 0.7%) and 4.6 + 0.01 mmol/1 in the control subjects (coefficient of variation: 5.1 +0.6%). Insulin-mediated glucose disposal was lower in the diabetic than in the non-diabetic subjects at the two lower insulin infusion rates (mean+ SEM=2.03+0.27 versus 4.8_+0.64mg-kg-l.min -~ at the first insulin infusion rate, p<0.01, and 5.59+0.59 versus 8.36_+0.61mg.kg-l.min -~ at the second insulin infusion rate, p < 0.01). However, insulin-induced glucose uptake did not differ significantly between the two groups at the third and fourth rates of insulin infusion. These results show that impaired insulin sensitivity in Type I diabetes is dependent on insulin concentration.Key words: Type 1 diabetes, glucose-controlled insulin infusion system, euglycaemic clamp, insulin dose response curve, free insulin, insulin antibodies.Many studies have shown that a large percentage of Type 2 (non-insulin-dependent) diabetic subjects are insulin resistant [1][2][3][4][5]. The situation in Type 1 (insulin-dependent) diabetes has not been clarified fully. At least four studies have been carried out to investigate insulin sensitivity in Type 1 diabetes using a fixed insulin infusion rate to generate steady-state plasma insulin levels. In one, where diabetes was of recent onset, nine out of 12 subjects had insulin sensitivity similar to that of controis [6]. In the other three studies [7][8][9], where the subjects had long-standing diabetes, insulin-mediated glucose utilization was markedly reduced in diabetic subjects, both under hyperglycaemic and euglycaemic conditions [8,9].The present study was undertaken to define more precisely the responsiveness to insulin in vivo in longstanding Type 1 diabetes. Studies in vitro have shown in many pathological conditions that abnormal tissue response to insulin is not seen over the full insulin dose response curve, but may be restricted to a relatively narrow range of insulin concentrations. We have therefore studied insulin sensitivity in vivo over a wide range of insulin concentrations. The method used was that validated by Rizza et al. [10] where four insulin infusion rates (corresponding to four widely different plasma insulin concentrations) were applied sequentially in four 2-h periods using the euglycaemic clamp technique.
Methods
SubjectsTwo groups were studied: (t) one group consisted of six male Type I diabetic patients, aged 23-45 years (mean 34 + 3 years) and 97%-123% of i...